Google Gemini and Bard artificial intelligence chatbot performance in ophthalmology knowledge assessment.

PURPOSE: With the popularization of ChatGPT (Open AI, San Francisco, California, United States) in recent months, understanding the potential of artificial intelligence (AI) chatbots in a medical context is important. Our study aims to evaluate Google Gemini and Bard's (Google, Mountain View, California, United States) knowledge in ophthalmology. METHODS: In this study, we evaluated Google Gemini and Bard's performance on EyeQuiz, a platform containing ophthalmology board certification examination practice questions, when used from the United States (US). Accuracy, response length, response time, and provision of explanations were evaluated. Subspecialty-specific performance was noted. A secondary analysis was conducted using Bard from Vietnam, and Gemini from Vietnam, Brazil, and the Netherlands. RESULTS: Overall, Google Gemini and Bard both had accuracies of 71% across 150 text-based multiple-choice questions. The secondary analysis revealed an accuracy of 67% using Bard from Vietnam, with 32 questions (21%) answered differently than when using Bard from the US. Moreover, the Vietnam version of Gemini achieved an accuracy of 74%, with 23 (15%) answered differently than the US version of Gemini. While the Brazil (68%) and Netherlands (65%) versions of Gemini performed slightly worse than the US version, differences in performance across the various country-specific versions of Bard and Gemini were not statistically significant. CONCLUSION: Google Gemini and Bard had an acceptable performance in responding to ophthalmology board examination practice questions. Subtle variability was noted in the performance of the chatbots across different countries. The chatbots also tended to provide a confident explanation even when providing an incorrect answer.

Preoperative Visual Acuity Thresholds in Pars Plana Vitrectomy for Epiretinal Membrane: A Systematic Review.

BACKGROUND AND OBJECTIVE: This review consolidates findings from studies that used a preoperative visual acuity (VA) threshold as an indication for epiretinal membrane (ERM) surgery. METHODS: The literature was systematically searched using Ovid MEDLINE, EMBASE, and Cochrane Library from January 2000 to October 2022 to select studies reporting on pars plana vitrectomy (PPV) for ERM that used a preoperative VA threshold as an inclusion criterion. Primary outcomes were final best-corrected visual acuity (BCVA) and change in BCVA relative to baseline. Secondary outcomes included risk of intra- and postoperative complications. RESULTS: A total of 639 eyes from seven studies were included. The most liberal preoperative VA threshold was 20/28.5 or worse, whereas the most conservative threshold was worse than 20/60. The mean preoperative BCVA was 0.55 logarithm of the minimum angle of resolution (logMAR) (∼20/70), and the mean postoperative BCVA was 0.35 logMAR (∼20/45). Generally, VA improved relative to baseline, regardless of the preoperative VA threshold. The smallest improvement in VA was observed in a study where the pre-operative VA to consider surgery was liberal (20/30 or worse), whereas the greatest VA improvement was observed in a study that used a conservative preoperative VA threshold (worse than 20/60). CONCLUSIONS: The greatest improvement in BCVA was observed in studies where a conservative pre-operative VA threshold was used. The decision to operate should involve a patient-centered approach with a thorough discussion of the risks and benefits of PPV, regardless of the preoperative VA threshold used. [Ophthalmic Surg Lasers Imaging Retina 2024;55:xx-xx.].

Accuracy of an Artificial Intelligence Chatbot's Interpretation of Clinical Ophthalmic Images.

Importance: Ophthalmology is reliant on effective interpretation of multimodal imaging to ensure diagnostic accuracy. The new ability of ChatGPT-4 (OpenAI) to interpret ophthalmic images has not yet been explored. Objective: To evaluate the performance of the novel release of an artificial intelligence chatbot that is capable of processing imaging data. Design, Setting, and Participants: This cross-sectional study used a publicly available dataset of ophthalmic cases from OCTCases, a medical education platform based out of the Department of Ophthalmology and Vision Sciences at the University of Toronto, with accompanying clinical multimodal imaging and multiple-choice questions. Across 137 available cases, 136 contained multiple-choice questions (99%). Exposures: The chatbot answered questions requiring multimodal input from October 16 to October 23, 2023. Main Outcomes and Measures: The primary outcome was the accuracy of the chatbot in answering multiple-choice questions pertaining to image recognition in ophthalmic cases, measured as the proportion of correct responses. χ2 Tests were conducted to compare the proportion of correct responses across different ophthalmic subspecialties. Results: A total of 429 multiple-choice questions from 136 ophthalmic cases and 448 images were included in the analysis. The chatbot answered 299 of multiple-choice questions correctly across all cases (70%). The chatbot's performance was better on retina questions than neuro-ophthalmology questions (77% vs 58%; difference = 18%; 95% CI, 7.5%-29.4%; χ21 = 11.4; P < .001). The chatbot achieved a better performance on nonimage-based questions compared with image-based questions (82% vs 65%; difference = 17%; 95% CI, 7.8%-25.1%; χ21 = 12.2; P < .001).The chatbot performed best on questions in the retina category (77% correct) and poorest in the neuro-ophthalmology category (58% correct). The chatbot demonstrated intermediate performance on questions from the ocular oncology (72% correct), pediatric ophthalmology (68% correct), uveitis (67% correct), and glaucoma (61% correct) categories. Conclusions and Relevance: In this study, the recent version of the chatbot accurately responded to approximately two-thirds of multiple-choice questions pertaining to ophthalmic cases based on imaging interpretation. The multimodal chatbot performed better on questions that did not rely on the interpretation of imaging modalities. As the use of multimodal chatbots becomes increasingly widespread, it is imperative to stress their appropriate integration within medical contexts.

Association Between Sociodemographic Factors and Self-reported Glaucoma in the National Health Interview Survey: A Population-Based Analysis.

PURPOSE: To investigate the association between social determinants of health (SDH) in the domains of social and community context, education access, environmental context, economic stability, and healthcare access, with glaucoma prevalence. DESIGN: Cross-sectional study. METHODS: The study population consisted of adult participants who answered glaucoma-related questions on the 2017 National Health Interview Survey (NHIS), the most recent iteration that includes glaucoma-related questions. The main outcome measures included the relationships between SDH-related factors and self-reported glaucoma diagnosis as well as self-reported glaucomatous vision loss were examined using univariable and multivariable regression models. RESULTS: In total, 26,696 of 26,742 (99.83%) NHIS respondents were included, of whom 880 (3.30%) reported a glaucoma diagnosis and 275 (1.03%) reported glaucomatous vision loss. Participants were predominantly middle-aged (50.95 ± 18.60 years), female (54.75%), and non-Hispanic White (70.49%). In age-adjusted multivariable regression (n = 25,456), non-Hispanic Black race (odds ratio [OR] = 1.87, 99% CI = [1.37, 2.55], P < .001, compared to non-Hispanic White race) and poor health status (OR = 1.54, 99% CI = [1.00, 2.37], P = .01, compared to good health status) were significant predictors of glaucoma diagnosis. For glaucomatous vision loss, having an income below the poverty threshold (OR = 2.41, 99% CI = [1.12, 5.20], P = .003, compared to income ≥5 times the poverty threshold) was the only significant predictor in univariable analyses. No SDH-related factors were significantly associated with glaucomatous vision loss in multivariable analysis (n = 848). Multicollinearity was minimal (variation inflation factor<1.6 for all independent variables). CONCLUSIONS: Non-Hispanic Black race and poor health status were associated with self-reported glaucoma diagnosis. Physicians and policymakers may consider SDH when assessing clinical risk and designing public health interventions.

PARS PLANA VITRECTOMY WITH OR WITHOUT INTERNAL LIMITING MEMBRANE PEEL FOR MACULAR HOLE: A Systematic Review and Meta-Analysis.

PURPOSE: To compare the efficacy and safety of pars plana vitrectomy with and without internal limiting membrane (ILM) peeling for macular hole (MH). METHODS: A systematic literature search on Ovid MEDLINE, Embase, Cochrane Library, and Google Scholar was performed from January 2000 to 2023. The primary outcome was the final best-corrected visual acuity (BCVA). Secondary outcomes included MH closure rates and the need for repeat surgery. The authors performed a random-effects meta-analysis on Review Manager 5.4. RESULTS: Fourteen studies on 880 eyes were included. Pars plana vitrectomy with and without ILM peel achieved a similar final BCVA ( P = 0.66). However, pars plana vitrectomy without ILM peeling achieved a significantly better final BCVA in eyes with closed MHs (WMD = 0.05 logMAR, 95% CI, 0.01-0.10, P = 0.02). Pars plana vitrectomy with ILM peeling achieved a significantly higher primary MH closure rate (RR = 1.21, 95% CI, 1.04-1.42, P = 0.02) and lower incidence of MH reoperation (RR = 0.19, 95% CI, 0.11-0.33, P < 0.001). The final MH closure rate ( P = 0.12) and incidence of MH recurrence ( P = 0.25) were similar between groups. CONCLUSION: Pars plana vitrectomy with and without ILM peel achieved a similar final BCVA. However, pars plana vitrectomy without ILM peeling achieved a better final BCVA in eyes with closed MHs. ILM peeling achieved a greater primary MH closure rate and reduced need for reoperation.

Artificial Intelligence Chatbots' Understanding of the Risks and Benefits of Computed Tomography and Magnetic Resonance Imaging Scenarios.

PURPOSE: Patients may seek online information to better understand medical imaging procedures. The purpose of this study was to assess the accuracy of information provided by 2 popular artificial intelligence (AI) chatbots pertaining to common imaging scenarios' risks, benefits, and alternatives. METHODS: Fourteen imaging-related scenarios pertaining to computed tomography (CT) or magnetic resonance imaging (MRI) were used. Factors including the use of intravenous contrast, the presence of renal disease, and whether the patient was pregnant were included in the analysis. For each scenario, 3 prompts for outlining the (1) risks, (2) benefits, and (3) alternative imaging choices or potential implications of not using contrast were inputted into ChatGPT and Bard. A grading rubric and a 5-point Likert scale was used by 2 independent reviewers to grade responses. Prompt variability and chatbot context dependency were also assessed. RESULTS: ChatGPT's performance was superior to Bard's in accurately responding to prompts per Likert grading (4.36 ± 0.63 vs 3.25 ± 1.03 seconds, P < .0001). There was substantial agreement between independent reviewer grading for ChatGPT (κ = 0.621) and Bard (κ = 0.684). Response text length was not statistically different between ChatGPT and Bard (2087 ± 256 characters vs 2162 ± 369 characters, P = .24). Response time was longer for ChatGPT (34 ± 2 vs 8 ± 1 seconds, P < .0001). CONCLUSIONS: ChatGPT performed superior to Bard at outlining risks, benefits, and alternatives to common imaging scenarios. Generally, context dependency and prompt variability did not change chatbot response content. Due to the lack of detailed scientific reasoning and inability to provide patient-specific information, both AI chatbots have limitations as a patient information resource.

The ability of artificial intelligence chatbots ChatGPT and Google Bard to accurately convey pre-operative information for patients undergoing ophthalmological surgeries.

INTRODUCTION: To determine whether the two popular artificial intelligence (AI) chatbots, ChatGPT and Bard, provide high-quality information concerning procedure description, risks, benefits, and alternatives of various ophthalmological surgeries. METHODS: ChatGPT and Bard were prompted with questions pertaining to the description, potential risks, benefits, alternatives, and implications of not proceeding with various surgeries in different subspecialties of ophthalmology. Six common ophthalmic procedures were included in our analysis. Two comprehensive ophthalmologists and one sub-specialist graded each response independently using a five-point Likert scale. RESULTS: Likert grading for accuracy was significantly higher for ChatGPT in comparison to Bard (4.5±0.6 vs 3.8±0.8, p<0.0001). Generally, ChatGPT performed better than Bard even when questions were stratified by type of ophthalmological surgery. There was no significant difference between ChatGPT and Bard for response length (2104.7±271.4 characters vs 2441.0±633.9 characters, p=0.12). ChatGPT responded significantly slower than Bard (46.0±3.0 seconds vs 6.6±1.2 seconds, p<0.0001). CONCLUSIONS: Both ChatGPT and Bard may offer accessible and high-quality information relevant to the informed consent process for various ophthalmic procedures. Nonetheless, both AI chatbots overlooked probability of adverse events, hence limiting their potential and introducing patients to information that may be difficult to interpret.

Anti-vascular endothelial growth factor therapy and retinal non-perfusion in diabetic retinopathy: A meta-analysis of randomised trials.

PURPOSE: Retinal non-perfusion (RNP) is fundamental to disease onset and progression in diabetic retinopathy (DR). Whether anti-vascular endothelial growth factor (anti-VEGF) therapy can modify RNP progression is unclear. This investigation quantified the impact of anti-VEGF therapy on RNP progression compared with laser or sham at 12 months. METHODS: A systematic review and meta-analysis of randomised controlled trials (RCTs) were performed; Ovid MEDLINE, EMBASE and CENTRAL were searched from inception to 4th March 2022. The change in any continuous measure of RNP at 12 months and 24 months was the primary and secondary outcomes, respectively. Outcomes were reported utilising standardised mean differences (SMD). The Cochrane Risk of Bias Tool version-2 and the Grading of Recommendations Assessment, Development and Evaluation (GRADE) guidelines informed risk of bias and certainty of evidence assessments. RESULTS: Six RCTs (1296 eyes) and three RCTs (1131 eyes) were included at 12 and 24 months, respectively. Meta-analysis demonstrated that RNP progression may be slowed with anti-VEGF therapy compared with laser/sham at 12 months (SMD: -0.17; 95% confidence interval [CI]: -0.29, -0.06; p = 0.003; I2 = 0; GRADE rating: LOW) and 24-months (SMD: -0.21; 95% CI: -0.37, -0.05; p = 0.009; I2 = 28%; GRADE rating: LOW). The certainty of evidence was downgraded due to indirectness and due to imprecision. CONCLUSION: Anti-VEGF treatment may slightly impact the pathophysiologic process of progressive RNP in DR. The dosing regimen and the absence of diabetic macular edema may impact this potential effect. Future trials are needed to increase the precision of the effect and inform the association between RNP progression and clinically important events. PROSPERO REGISTRATION: CRD42022314418.

Pars Plana Vitrectomy with or without Internal Limiting Membrane Peel for Epiretinal Membrane: A Systematic Review and Meta-Analysis.

BACKGROUND: The safety and effectiveness of pars plana vitrectomy (PPV) with internal limiting membrane (ILM) peeling remains poorly understood for the treatment of epiretinal membrane (ERM). OBJECTIVES: Our study aims to compare the safety and effectiveness of PPV with and without ILM peeling for ERM. METHODS: A systematic literature search was conducted on Ovid MEDLINE, Embase, Cochrane Library, and Google Scholar from January 2000 to January 2023 for comparative studies reporting visual and anatomical outcomes for patients with ERM that received PPV with or without ILM peeling. Primary outcomes included best-corrected visual acuity (BCVA) at last study observation and change in BCVA from baseline. Secondary outcomes included retinal thickness (RT) at last study observation, change in RT from baseline, risk of ERM recurrence, and adverse events. A random-effects meta-analysis was performed. Risk of bias of randomized controlled trials was assessed using the Risk of Bias 2 tool of observational studies using the Risk of Bias in Non-randomized Studies of Interventions-I tool. The certainty of evidence of outcomes was evaluated using Grading of Recommendations, Assessment, Development and Evaluation criteria. RESULTS: Nineteen studies reporting on 1,291 eyes at baseline were included. PPV with and without ILM peel achieved a similar BCVA at last study observation (p = 0.68) and change in BCVA from baseline (p = 0.79). These findings remained consistent irrespective of whether simultaneous phacoemulsification was performed. PPV with ILM peel achieved a significantly lower incidence of ERM recurrence (risk ratio [RR] = 0.26, 95% CI = [0.13, 0.51], p < 0.0001) and additional surgery (RR = 0.17, 95% CI = [0.04, 0.74], p = 0.02) compared to PPV without ILM peel. CONCLUSION: PPV with and without ILM peel achieved a similar BCVA at last study observation in ERM patients. Patients treated with PPV and ILM peel also had a reduced risk of ERM recurrence and lower reoperation risk. These conclusions are associated with a moderate certainty of evidence and potential for bias from multiple non-randomized studies.

Intravitreal Triamcinolone Acetonide for Diabetic Macular Edema and Macular Edema Secondary to Retinal Vein Occlusion: A Meta-Analysis.

BACKGROUND: The comparative safety and efficacy of different doses of intravitreal triamcinolone acetonide (IVTA) for diabetic macular edema (DME) and macular edema (ME) secondary to retinal vein occlusion (RVO) is unclear. OBJECTIVES: This meta-analysis aimed to compare the safety and efficacy of different doses of IVTA in this setting. METHODS: A systematic literature search for randomized clinical trials (RCTs) was conducted on Cochrane Library, Ovid MEDLINE, and EMBASE from January 2005 to May 2022. Studies that reported on patients with DME or ME secondary to RVO that received treatment with different doses of IVTA were included. A random-effects meta-analysis was performed. Cochrane's Risk of Bias Tool 2 was used to assess the risk of bias, and Grading of Recommendations, Assessment, Development and Evaluation (GRADE) guidelines were used to assess certainty of evidence. RESULTS: Five RCTs reporting on 1,041 eyes at baseline were included in this meta-analysis. In eyes with ME secondary to RVO, high-dose (4 mg) IVTA achieved a significantly better change in best-corrected visual acuity (WMD = -4.75 ETDRS letters, 95% CI = [-7.73, -1.78], p = 0.002) and reduction in retinal thickness (WMD = -93.02 µm, 95% CI = [-153.23, -32.82], p = 0.002) at months 4-6 compared to low-dose (1-2 mg) IVTA. However, high-dose IVTA had a higher risk of intraocular pressure-related adverse events (RR = 2.99, 95% CI = [1.05, 8.50], p = 0.04) and cataract surgery (RR = 5.67, 95% CI = [3.09, 10.41], p < 0.00001) than low-dose IVTA in eyes with ME secondary to RVO. These efficacy and safety differences in high-dose and low-dose IVTA were not observed in DME eyes. CONCLUSIONS: The RCT evidence in this setting is limited. High-dose IVTA achieved greater improvements in visual acuity and reductions in retinal thickness than low-dose IVTA at months 4-6. However, high-dose IVTA had a less favorable safety profile than low-dose IVTA. The significance of these outcomes was based on patients with ME secondary to RVO, but not DME.

Histological changes in retinal detachment: A systematic review for the clinician.

Although there have been numerous innovations in the management of retinal detachment (RD) over the past decades, there is still limited understanding of the pathophysiological processes that take place before and after repair. Summarizing key concepts using animal studies may allow for a better assessment of common pre- and postoperative microstructural abnormalities in RD. We performed a systematic literature review on Ovid MEDLINE, EMBASE, and Cochrane Controlled Register of Trials from January 1968 to January 2022, searching animal or human studies reporting retinal histologic changes following primary or induced RD. Thirty-two studies were included. Main cellular events were summarized: photoceptor apoptosis occurs as early as 12 hours after RD and, although most cells survive, there is extensive remodeling. Outer segments progressively degenerate, while inner segments are reorganized. Rod and cone opsins are redistributed, and rod axons retract while cones undergo changes in shape. Second- and third-order neurons rearrange their dendritic processes, and Müller cells become hypertrophic, growing into the subretinal space. Finally, retinal pigment epithelium cells undergo a change in their morphology. Acknowledging critical morphologic changes following RD is crucial in understanding why anatomical and functional outcomes can vary. Insights from histological studies, together with high-resolution imaging, may be key in identifying novel biomarkers in RD.

Association Between Vision Difficulty and Sociodemographic Factors in Children: A Population-Based Analysis.

PURPOSE: To determine sociodemographic predictors of poor vision in children and adolescents less than 18 years of age in a large, nationally representative sample of the US population. DESIGN: Retrospective, population-based cross-sectional study. METHODS: Using data from the 2021 National Health Interview Survey (NHIS), participants less than 18 years of age for whom data were available on vision difficulty were included in our analysis. Our primary outcome was vision difficulty. Sociodemographic variables were investigated as predictors of vision difficulty. Logistic regression models were performed using Stata version 17.0. An odds ratio (OR) and 95% CI were reported for analyses. RESULTS: The 2021 NHIS reported on 8261 children and adolescents, of whom 7373 had data pertaining to vision difficulty and were included in our sample. Vision difficulty was associated with being unable to afford medical care (OR = 2.60, 95% CI = 1.17, 5.80, P = 0.02) and having public health insurance (OR = 1.52, 95% CI = 1.09, 2.12, P = .01). Compared to children less than 5 years of age, children ranging from 5 to 8 years (OR = 4.29, 95% CI = 2.26, 8.15, P < .01) and adolescents ranging from 16 to 17 years (OR = 6.06, 95% CI = 3.17, 11.58, P < .01) had a higher odds of vision difficulty. Compared to participants for whom the highest level of education of adults in their family was grade 1 to 11, the odds of vision difficulty were lower in those whose parents had a professional school or doctoral degree (OR = 0.30, 95% CI = 0.11, 0.84, P = .02). CONCLUSIONS: Multiple sociodemographic factors are associated with vision difficulty in young persons. In working toward achieving equity in vision health, it is imperative that disparities mediated by sociodemographic factors be addressed through public health policies.

Periorbital Necrotizing Fasciitis: Case Presentation.

Necrotizing fasciitis (NF) is an aggressive and potentially life-threatening infection of the superficial fascia and surrounding skin, fat, fascia, muscle, and other soft tissue structures. Here, we outline the rare case of a 26-year-old man with a periorbital Streptococcus pyogenes A NF infection. Our case report underscores a unique instance of periorbital NF, distinctively presenting without any predisposing risk factors, shedding light on its presentation, treatment, and pathophysiology.

Understanding the diagnosis of superior mesenteric artery syndrome: analysis of the location of duodenal impression on upper gastrointestinal studies.

BACKGROUND: Upper gastrointestinal (GI) contrast studies are frequently requested to aid superior mesenteric artery syndrome diagnosis, a rare entity. Compression of the third duodenal part is expected to be mid-to-left of the midline where the superior mesenteric artery arises from the aorta; however, a duodenal impression to the right of the midline due to normal anatomic impression by the inferior vena cava (IVC) is often encountered and frequently misdiagnosed. OBJECTIVE: The purpose of this study was to determine the frequencies of (1) normal right-of-midline duodenal impressions and (2) mid-to-left of midline compressions in upper GI studies in a tertiary pediatric referral center. MATERIALS AND METHODS: All upper GI studies performed at our institution over 2 years were retrospectively evaluated to determine whether the duodenum had vertical duodenal impression to the right of the vertebral midline, mid-to-left of the vertebral midline, or no identifiable duodenal impression at all. RESULTS: In total, 538 upper GI studies were included in this analysis. A total of 275 male and 247 female patients between 0 and 17 years of age (median: 6 years, range: 1 month-17 years) were included. Of 538 total upper GI studies, there were 240 studies (44.6%) with a right-of-midline impression. There were only 10 studies (1.9%) with a mid-to-left of midline compression, and 9/10 also showed a concurrent right-sided impression sign. CONCLUSION: Right-of-midline duodenal impression is a normal anatomic finding caused by the IVC and should not be confused with superior mesenteric artery syndrome. In the presence of an appropriate clinical context, proximal duodenal dilation, "to-and-fro" motion of contrast, and duodenal impression at mid-to-left of midline, a diagnosis of superior mesenteric artery syndrome should be considered.